Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Chapter 15: Pricing and the Revenue Management utdallas.edu/~metin 1 Outline The Role of RM (Revenue Management) in the SCs RM for Multiple Customer Segments RM for Perishable Assets RM for Seasonable Demand RM for Bulk and Spot Customers Using RM in Practice Summary of Learning Objectives 2 utdallas.edu/~metin The Role of RM in SCs Revenue management is the use of pricing to increase the profit generated from a limited supply of supply chain assets – SCs are about matching demand and capacity – Prices affect demands Yield management similar to RM but deals more with quantities rather than prices Supply assets exist in two forms – Capacity: expiring – Inventory: often preserved Revenue management may also be defined as offering different prices based on customer segment, time of use and product or capacity availability to increase supply chain profits Most common example is probably in airline ticket pricing – Pricing according to customer segmentation at any time – Pricing according to reading days for any customer segment » Reading days: Number of days until departure utdallas.edu/~metin 3 Conditions for RM to Work The value of the product varies in different market segments – Airline seats: Leisure vs. Business travel – Films: Movie theater goers, DVD buyers, Cheap movie theater goers, TV watchers. The product is highly perishable or product waste occurs – Fashion and seasonal apparel – High tech products Demand has seasonal and other peaks – Products ordered at Amazon.com, peaking in December – Supply Chain textbook orders peaking in August and January. The product is sold both in bulk and on the spot market – Owner of warehouse who can decide whether to lease the entire warehouse through long-term contracts or save a portion of the warehouse for use in the spot market – Truck capacities for a transportation company 4 utdallas.edu/~metin RM for Multiple Customer Segments If a supplier serves multiple customer segments with a fixed asset, the supplier can improve revenues by setting different prices for each segment – Must figure out customer segments Prices must be set with barriers such that the segment willing to pay more is not able to pay the lower price – Barriers: Time, location, prestige, inconvenience, extra service In the case of time barrier, – The amount of the asset reserved for the higher price segment is such that quantities below are equal » the expected marginal revenue from the higher priced segment » the price of the lower price segment 5 utdallas.edu/~metin Barrier: Extra service Price What is the targeted demand? What is the Revenue? Could we do better? Money left on the table =160,000 P0=1200 Revenue=480,000 P=2000-2Q C=400 utdallas.edu/~metin No. seats 6 Customer Segmentation by extra service Example: Cruise ship A cruise ship with C=400 identical cabins What is the price to maximize revenue? 2000 Price P=2000-2Q Demand Curve 1000 No. Seats utdallas.edu/~metin 7 Example: Cruise ship Offer additional services to differentiate products and pricing Price Revenue=1600(200) + 1200(400-200)=560,000 P2=1600 Increase revenue more? P1=1200 Q2=200 utdallas.edu/~metin Q1 =400 No. seats 8 Example Price P3=1800 Revenue=1800(100) + 1600(200-100) + 1200(400-200)=580,000 P2=1600 P1=1200 How to allocate capacity for each product/service optimally? Q3=100 Q2=200 utdallas.edu/~metin Q1 =400 No. seats 9 Barrier: Time which implies Customer Segment RM for Multiple Customer Segments pL = the price charged to the lower price segment pH = the price charged to the higher price segment DH = mean demand for the higher price segment σH = standard deviation of demand for the higher price segment CH = capacity reserved for the higher price segment RH(CH) = expected marginal revenue from reserving more capacity = Prob(Demand from higher price segment > CH) x pH Optimality by equivalance of marginal revenues: RH(CH) = pL which leads to Prob(Demand from higher price segment > CH) = pL / pH CH = F-1(1- pL/pH, DH,σH) = Norminv(1- pL/pH, DH,σH) utdallas.edu/~metin 10 Example 15.1: ToFrom Trucking Revenue from segment A = pA = $3.50 per cubic ft Revenue from segment B = pB = $2.00 per cubic ft Mean demand for segment A = DA = 3,000 cubic ft Std dev of segment A demand = σA = 1,000 cubic ft = Norminv(1- pB/pA, DA,σA) CA = Norminv(1- (2.00/3.50), 3000, 1000) = 2,820 cubic ft If pA increases to $5.00 per cubic foot, then CA = Norminv(1- pB/pA, DA,σA) = Norminv(1- (2.00/5.00), 3000, 1000) utdallas.edu/~metin = 3,253 cubic ft 11 Two questions: What happens to the capacity reserved for the high paying segment, when the high paying segment starts paying about the low paying segment? – How much does the high paying segment value quick service? We never consider the distribution of the demand for low paying segment when computing the reserved capacity for the high paying segment. Can this be correct, why? 12 utdallas.edu/~metin RM in the Service Industries Airline Industry uses RM the most. Evidence of airline revenue increases of 4 to 6 percent: – With effectively no increase in flight operating costs RM allows for tactical matching of demand vs. supply: – Booking limits can direct low-fare demand to empty flights – Protect seats for highest fare passengers on forecast full flights Hotel, Restaurant, Car rental, Overseas shipping, Cruise travel, Transportation capacity providers, Computation capacity providers (computer farms) and sometimes Health care industries show similarities to airline industry in using RM. 13 utdallas.edu/~metin What are the barriers among customer segments in the airline industry? Sensitivity to Duration Sensitivity to Flexibility Low High Leisure No Travelers Demand No Business Offer Travelers Sensitivity to Price High utdallas.edu/~metin Low 14 A RM Model Expected Revenue = Expected Revenue from Business Class + Expected Revenue from Leisure Class Assume demand distribution for the business class is known and the demand for leisure class is ≥ C rB = revenue/business passenger, rL = revenue/leisure passenger, C = Airplane capacity ∞ Q R(Q ) = rB ∫ xf ( x )dx + Q ∫ f ( x )dx + (C − Q )rL 0 Q dR(Q ) = rB [1 − F (Q )] − rL = 0 dQ Marginal revenue of business class = Marginal revenue of leisure class rB − rL F (Q ) = rB * utdallas.edu/~metin SL: Service Level 15 Example: Airline seat classes There are only two price classes – Leisure: (f2) $100 per ticket – Business: (f1) $250 per ticket Total available capacity= 80 seats Distribution of demand for business class is known Assume enough demand for the leisure class How many seats to allocate to the business class to maximize expected revenue? 16 utdallas.edu/~metin Business Class Demand Distribution 100% Probability 80% 60% 40% 20% 0% 0 5 10 15 20 25 30 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Probability cdf Probability 5% 11% 28% 22% 18% 10% 6% cdf 5% 16% 44% 66% 84% 94% 100 Business Class Seat Allocation 17 utdallas.edu/~metin Identifying Optimal Allocation 100% SL= 0.60 Probability 80% 60% 40% 20% Note: rB: 250, rL: 100 SL = (rB - rL )/ rB = (250-100)/250 = 60% 0% 0 5 10 15 20 25 Probability cdf Probability 5% 11% 28% 22% 18% 10% 6% cdf 5% 16% 44% 66% 84% 94% 100 Business Class Seat Allocation Optimal Allocation Quantity = 15 utdallas.edu/~metin 30 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 18 Expected Revenue Expected Revenue 9600 9400 9200 9000 8800 8600 8400 8200 8000 7800 9438 9363 9238 9063 8688 8638 8000 0 5 10 15 20 25 30 35 Business Class 19 utdallas.edu/~metin Optimality Condition Optimality Condition: Choose the number of seats for the business class such that marginal revenue from business class is the same as the marginal revenue from the leisure class. Marginal Revenue Business 250 200 150 Marginal Revenue Leisure 100 50 0 0 5 10 15 20 25 30 35 20 utdallas.edu/~metin RM for Perishable Assets Any asset that loses value over time is perishable Examples: high-tech products such as computers and cell phones, high fashion apparel, underutilized capacity, fruits and vegetables Two basic approaches: – Dynamic Pricing: Vary price over time to maximize expected revenue – Overbooking: Overbook sales of the asset to account for cancellations » » » » Airlines use the overbooking most Passengers are “offloaded” to other routes Offloaded passengers are given flight coupons This practice is legal – Dynamic pricing belongs to RM while overbooking can be said to more within the domain of Yield management. » But concepts are more important than the names! 21 utdallas.edu/~metin RM for Perishable Assets Overbooking or overselling of a supply chain asset is valuable if order cancellations occur and the asset is perishable The level of overbooking is based on the trade-off between the cost of wasting the asset if too many cancellations lead to unused assets (spoilage) and the cost of arranging a backup (offload) if too few cancellations lead to committed orders being larger than the available capacity Spoilage and offload are actually terms used in the airline industry 22 utdallas.edu/~metin RM for Perishable Assets p = price at which each unit of the asset is sold c = cost of using or producing each unit of the asset b = cost per unit at which a backup can be used in the case of shortage Cw = p – c = marginal cost of wasted capacity = Overage cost Cs = b – c = marginal cost of a capacity shortage = Underage cost O* = optimal overbooking level P(Demand<Capacity)= Cu / (Cu + Co) P(Demand>=Capacity)= Co / (Cu + Co) P(Order cancellations< O*) = Co / (Cu + Co) s* := Probability(order cancellations < O*) = Cw / (Cw + Cs) Beware: This is the newsvendor formula in disguise. utdallas.edu/~metin 23 RM for Perishable Assets If the distribution of cancellations is known to be normal with mean µc and standard deviation σc then O* = F-1(s*, µc, σc) = Norminv(s*, µc, σc) If the distribution of cancellations is known only as a function of the booking level (capacity L + overbooking O) to have a mean of µ(L+O) and std deviation of σ(L+O), the optimal overbooking level is the solution to the following equation: O * = F-1(s*,µ(L+O),σ(L+O)) = Norminv(s*,µ(L+O),σ(L+O)) 24 utdallas.edu/~metin Example 15.2 Cost of wasted capacity = Cw = $10 per dress Cost of capacity shortage = Cs = $5 per dress s* = Cw / (Cw + Cs) = 10/(10+5) = 0.667 µc = 800; σc = 400 O* = Norminv(s*, µc,σc) = Norminv(0.667,800,400) = 973 If the mean is 15% of the booking level and the coefficient of variation is 0.5, then the optimal overbooking level is the solution of the following equation: O * = Norminv(0.667,0.15(5000+O * ),0.075(5000+O * )) Using Excel Solver, O* = 1,115 25 utdallas.edu/~metin RM for Seasonal Demand Seasonal peaks of demand are common in many SCs\ – Most retailers achieve a large portion of total annual demand in December » Amazon.com Off-peak discounting can shift demand from peak to nonpeak periods Charge higher price during peak periods and a lower price during off-peak periods Read Section 9.3: Managing Demand [with discounts] of the textbook. 26 utdallas.edu/~metin RM for Bulk and Spot Customers Most consumers of production, warehousing, and transportation assets in a supply chain face the problem of constructing a portfolio of long-term bulk contracts and short-term spot market contracts – Long-term contracts for low cost – Short-term contracts for flexibility The basic decision is the size of the bulk contract The fundamental trade-off is between wasting a portion of the lowcost bulk contract and paying more for the asset on the spot market 27 utdallas.edu/~metin RM for Bulk and Spot Customers For the simple case where the spot market price is known but demand is uncertain, a formula can be used cB = bulk rate cS = spot market price Q* = optimal amount of the asset to be purchased in bulk p* = probability that the demand for the asset does not exceed Q* Marginal cost of purchasing another unit in bulk is cB. The expected marginal cost of not purchasing another unit in bulk and then purchasing it in the spot market is (1-p*)cS. 28 utdallas.edu/~metin Revenue Management for Bulk and Spot Customers If the optimal amount of the asset is purchased in bulk, the marginal cost of the bulk purchase should equal the expected marginal cost of the spot market purchase, or cB = (1-p*)cS Solving for p* yields p* = (cS – cB) / cS If demand is normal with mean µ and std deviation σ, the optimal amount Q* to be purchased in bulk is Q* = F-1(p*,µ,σ) = Norminv(p*,µ,σ) 29 utdallas.edu/~metin Example 15.3: Buying transportation capacity to bring goods from China Bulk contract cost = cB = $10,000 per million units Spot market cost = cS = $12,500 per million units Demand for transportation: µ = 10 million units σ = 4 million units p* = (cS – cB) / cS = (12,500 – 10,000) / 12,500 = 0.2 Q* = Norminv(p*,µ,σ) = Norminv(0.2,10,4) = 6.63 The manufacturer should sign a long-term bulk contract for 6.63 million units per month and purchase any transportation capacity beyond that on the spot market If the demand is exactly 10 million units without any variability, how much long-term bulk contract with the transporter? 30 utdallas.edu/~metin Using RM in Practice Evaluate your market carefully – – – – Understand customer requirements for services and products Price, flexibility (time, specs), value-added services, etc. Based on requirements identify customer segments (groups) Differentiate products/services and their pricing according to customer segments » Dell: » Same product is sold at a different price to different consumers (private/small or large business/government/academia/health care) » Price of the same product for the same industry varies 31 utdallas.edu/~metin Using RM in Practice Quantify the benefits of revenue management Implement a forecasting process Apply optimization to obtain the revenue management decision Involve both sales and operations Understand and inform the customer Integrate supply planning with revenue management 32 utdallas.edu/~metin Summary of Learning Objectives What is the role of revenue management in a supply chain? Under what conditions are revenue management tactics effective? What are the trade-offs that must be considered when making revenue management decisions? 33 utdallas.edu/~metin Smart Pricing Through Rebates utdallas.edu/~metin 34 Rebate Examples Nikon Coolpix digital camera is sold either on-line or in stores for $600. the manufacturer provides a rebate of $100 independently of where the camera is purchased. Sharp VL-WD255U digital camcorder is sold for about $500 at retail or virtual stores. Sharp provides a rebate to the customer of $100 independently of where the product is purchased. … utdallas.edu/~metin 35 36 utdallas.edu/~metin 37 utdallas.edu/~metin Mail-in-Rebate What is the manufacturer trying to achieve with the rebate? – Why the manufacturer and not the retailer? Should the manufacturer reduce the wholesale price instead of the rebate? Are there other strategies that can be used to achieve the same effect? 38 utdallas.edu/~metin Example 10000 A Retailer and a manufacturer. Demand Curve Demand P=2000-0.22Q – Retailer faces customer demand. – Retailer orders from manufacturer. 2000 Price Variable Production Cost=$200 Selling Price=? Retailer Manufacturer Wholesale Price=$900 utdallas.edu/~metin 39 Example Retailer profit=(PR-PM)(1/0.22)(2,000 - PR) Manufacturer Retailer profit=(PM-CM) (1/0.22)(2,000 - PR) takes PM=$900 – Sets PR=$1450 to maximize (PR -900) (1/0.22)(2,000 - PR) – Q = (1/0.22)(2,000 – 1,450) = 2,500 units – Retailer Profit = (1,450-900)·2,500 = $1,375,000 Manufacturer takes CM=variable cost – Manufacturer profit=(900-200)·2,500 = $1,750,000 utdallas.edu/~metin 40 Retailer Expected Profit (No Rebate) 1,600,000 $1,375,000 1,400,000 Retailer Expected Profit 1,200,000 1,000,000 800,000 600,000 400,000 200,000 0 500 1,000 1,500 2,000 2,500 3,000 3,500 3,654 4,110 4,567 4,547 Order 41 utdallas.edu/~metin Manufacturer Profit (No Rebate) 6,000,000 Manufacturer Profit 5,000,000 4,000,000 3,000,000 $1,750,000 2,000,000 1,000,000 55 7, 8 28 41 7, 4 7, 0 14 6, 6 01 6, 2 74 88 5, 7 5, 3 61 4, 9 47 4, 5 10 67 4, 5 4, 1 54 3, 6 00 00 3, 5 3, 0 00 2, 5 00 2, 0 00 00 1, 5 1, 0 50 0 0 Order 42 utdallas.edu/~metin Example: Customer Mail-in Rebate What happens with $100 customer mail-in rebate? – Note that it is a discount for the customer so the demand should go up!!! » Q = (1/0.22) [2,000 – (PR-Rebate)] = (1/0.22) [2,000 – (1450-100)] = 2954 – Retailer Profit = (1,450-900)·2,955 = $1,625,250 – Manufacturer profit=(900-200-100)·2,955 = $1,773,000 43 utdallas.edu/~metin Retailer Expected Profit ($100 Rebate) $1,625,250 1,800,000 1,600,000 1,400,000 Retailer Expected Profit 1,200,000 1,000,000 800,000 600,000 400,000 200,000 0 1,000 1,500 2,000 2,500 3,000 3,500 Order utdallas.edu/~metin 4,000 4,110 4,567 4,547 4,961 44 Manufacturer Profit ($100 Rebate) 6,000,000 Manufacturer Profit 5,000,000 4,000,000 3,000,000 2,000,000 $1,773,000 1,000,000 1, 00 0 1, 50 0 2, 00 0 2, 50 0 3, 00 0 3, 50 0 4, 00 0 4, 11 0 4, 56 7 4, 54 7 4, 96 1 5, 37 4 5, 78 8 6, 20 1 6, 61 4 7, 02 8 7, 44 1 7, 85 5 8, 26 8 0 Order utdallas.edu/~metin 45 Example: Wholesale discount What happens with $100 wholesale discount to retailer? Retailer takes PM=$800 – Sets PR=$1400 to maximize (PR -800) (1/0.22)(2,000 - PR) – Q = (1/0.22)(2,000 – 1,400) = 2,727 units – Retailer Profit = (1,400-800)·2,727 = $1,499,850 Manufacturer takes PR=$800 and CM=variable cost – Manufacturer profit=(800-200)·2,727 = $1,499,850 46 utdallas.edu/~metin Example: Global Optimization What happens if manufacturer sells directly or optimizes for the whole system globally? – Manufacturer sets PR=$1,100 to maximize (PM -200) (1/0.22)(2,000 - PM) – Q = (1/0.22)(2,000 – 1,100) = 4,091 units – Manufacturer profit=(1100-200)· 4,091 = $3,681,900 47 utdallas.edu/~metin Strategy Comparison Strategy No Rebate With Rebate ($100) Reduce Wholesale P ($100) Global Optimization Retailer 1,370,096 1,625,250 1,499,850 Manufacturer 1,750,000 1,773,000 1,499,850 Total 3,120,096 3,398,250 2,999,700 3,681,900 48 utdallas.edu/~metin Managerial Insights Mail in Rebate allows supply chain partners to move away from sequential strategies toward global optimization – Provides retailers with upside incentive Mail in Rebate outperforms wholesale price discount for manufacturer Other advantages of rebates: – Not all customers will remember to mail them in – Gives manufacturer better control of pricing utdallas.edu/~metin 49 Smart Pricing Customized Pricing – Revenue Management Techniques » Distinguish between customers according to their price sensitivity – Influence retailer pricing strategies – Move supply chain partners toward global optimization Dynamic Pricing – Changing prices over time without necessarily distinguishing between different customers – Find the optimal trade-off between high price and low demand versus low price and high demand 50 utdallas.edu/~metin When does Dynamic Pricing Provide Significant Profit Benefit? Limited Capacity Demand Variability Seasonality in Demand Pattern Short Planning Horizon 51 utdallas.edu/~metin The Internet makes Smart Pricing Possible Low Menu Cost Low Buyer Search Cost Visibility – To the back-end of the supply chain allows to coordinate pricing, production and distribution Customer Segmentation – Difficult in conventional stores and easier on the Internet Testing Capability 52 utdallas.edu/~metin A Word of Caution Amazon.com experimented with dynamic pricing – customers responded negatively Coca-Cola distributors rebelled against a seasonal pricing scheme Opaque fares (priceline.com, hotwire.com) – Determining the correct mix of opaque and regular fares is difficult. 53 utdallas.edu/~metin